| 摘要 |
A reasoning system includes at least one atomic context and a plurality of derivation methods. Each atomic context encapsulates a reference to at least one mathematical model and is configured to cause the performance of at least one computation using the encapsulated mathematical model. Each of the derivation methods is configured to receive a reference to a base context and to create a derived context that is based on the base context. Such created derived contexts are configured to cause the performance of a computation by invoking the base context and providing at least one input to the base context, to receive an output from the base context, to perform a transformation on an input to the base context or on an output from the base context, and to return at least one output of the computation. Derived contexts can be composed in order to create more complex contexts. |
